I love cursing more than most people, but it doesn't belong in my code.

I agree. I love cursing, but I think it works best in more ephemeral situations. In conversation (spoken, written, or otherwise) it works great but in official records that will be often revisited (whether they are source code comments or commit messages) the cursing will eventually come off as tacky as the emotional context that warranted it is lost in time.

If I am browsing through source from 10 years ago, it may help me if a comment alerts me that "this fucking cockstain of a function broke prod", but really there is no connect between me and the emotion that warranted the profanity. I need more context, otherwise it seems tacky.

(I think that recorded conversation, although not ephemeral, retains the emotional context necessary for profanity to 'work')

that is exactly right

it would also get one in trouble at many companies. Anyone's that's ever been through an acquisition will know that due diligence includes code scans - it's highly unprofessional to let cursing into a code base.

I am confused by the suggestion that the word "love" would indicate happiness. Usually when I use that in source code, I'm being sarcastic. As in, Don't you love the fact that Google's API documentation is inaccurate?

This is probably a homage to the original Linux Kernel Fuck Count[1] which included love in its search.

http://durak.org/sean/pubs/kfc/

Yes, but statistically that will be mere noise, just like I love love love this algorithm or this algorithm works hand in g love with that other one.

Maybe. It is possible that the word is used sarcastically more often than literally though. For instance, it wouldn't surprise me to learn that the word "literally" may actually be used more often to mean "not literally".

'Literally' is already in the OED as 'not literally': http://www.telegraph.co.uk/education/10240917/Uproar-as-OED-...

Indeed! http://lee-phillips.org/literallyEgregious/

k, brb, building a git sarcastic tool

I like this, I'd probably use it to track the number of TODO's that are added to the code over time.

I've found out why they build up — it seems that the elves have been ignoring them for years now. Makes no sense, I know, but they've gone back to making shoes overnight. It doesn't matter how many times I write TODO, it never actually gets done unless I do it myself anymore.

This reminds me of http://commitlogsfromlastnight.com/

Looks like you're probably running it with Ruby 1.8, and it's using the Ruby 1.9 Hash syntax. Try to run it with 1.9, either through something like RVM, rbenv or chruby, or with any other number of distribution installers.

Having the rate of change rather than the number of words might be a better indicator of how stressed a particular group are at a given time.